I am working with a dataset with daily frequency with a total of 1461 observations.
I performed Unit Root Tests to my variables but I ended up with some different results, as you can see in the image below:
- urt.png (44.81 KiB) Viewed 6728 times
I have painted in yellow the results that tell me that my data is I(1) with a significance level of 5%. As you can see, with PP test all my variables are I(0) except when I remove both trend and constant.
However, if I look at the results of the ADF test, I see some different results in the variables "imp" "Hydric" and "Price", suggesting that those variables are I(1).
I have also performed the KPSS test, and it tells me even more different results, since it has suggested me that all my variables are I(1).
In a case like this, what is the most accurate thing to conclude regarding the stationary level of my variables? Are they I(0) or I(1)?
As a further note, I will be most likely use the GARCH and\or VAR model. Also, all the variable are in their natural logs and I have removed the outliers.
I appreciate all the help in advance.
Yours sincerely,
Afonso Rodrigues